Ph8ddlZddlZddlZddlZddlZddlmZddlmZddl m Z m Z ddl Z ddl Z ddl mZddlmZddlmZddlmZddlmZdd lmZmZmZdd lmZdd lmZmZm Z ddl!m"cm#Z$ejJe&Z'd Z(ed Z)edejTde fdZ+d dZ,dZ-edejTde fdZ.Gddej^Z0edejTde fdZ1edZ2dZ3e jhjjZ5e5jle5jne5jpe5jre5jte5jve5jxe5jze5j|e5j~e5je5je5je5jje5jje5je5je5je5je5je5je5jhZMeeMZMedZNde e ejffdZPdZQdaRdZSdZTd!dZUy)"N)contextmanager)partial)CallableUnion)SymInt)get_decompositions) bind_symbols) aot_function aot_modulemake_boxed_compiler)strip_overloads)default_partition draw_graph#min_cut_rematerialization_partitionc|jjD]]}|jtjj j k(s5tjj j|__|j|SN) graphnodestargettorchopsaten_to_copyto recompile)fx_gnodes eC:\Users\daisl\Desktop\realtime-object-detection\venv\Lib\site-packages\torch/_functorch/compilers.py _canonicalizer !sV    ;;%))..11 1))..++DK! NN Kc#Ktjjd} dtjj|y#tjj|wxYww)NF)r_C_jit_set_autocast_mode)old_jit_autocast_flags r_disable_jit_autocastr&)sI!HH;;EB?  ''(=>''(=>s A+A A+!A((A+rreturncVt5t||jjD]}|jt j jjk(s5t|jdk(sNt|jdk(sgd|jvsvt j jj|_|jjD]X}i}|jjD]0\}}t|t jr |j }|||<2||_ Z|jj#|j%t j&j)|}t j*j-|jt j&j/|j1}t j&j3|}t5d|Ds||ddd|S#1swYSxYw)a Compiles the :attr:`fx_g` with Torchscript compiler. .. warning:: This API is experimental and likely to change. Args: fx_g(fx.GraphModule): The input Fx graph module to be compiled. Returns: Torch scripted model. r dtypec3dK|](}t|tjj*ywr) isinstancer _subclasses FakeTensor).0ts r zts_compile.._s$M1:a!2!2!=!=>s.0N)r&rrrrrrrrlenargskwargsritemsr+devicetypelintrjitscriptr#_jit_pass_remove_mutationfreezeevaloptimize_for_inferenceany)rinpsr new_kwargskvfs r ts_compilerD2s  JJ$$D uyy~~666 Na' $)t{{*#iinn// %JJ$$DJ ))+1a.A ! 1 ,%DK %   II  T " **1773 II  QVVX & II , ,Q /MMM tH? !@ HA !@ Hs%AH"H;HH#E1HH(cLt|jt||||S)N) clear_meta)printcoder)r_namerFs r_draw_graph_compilerKds $))tTj1 Kr!c6ttt|S)NrJ)r rrKrMs rdraw_graph_compilerNjs #$/ r!c|S)z Returns the :attr:`fx_g` Fx graph module as it is. This is a no-op compiler and can be used to check accuracy. .. warning:: This API is experimental and likely to change. rrIs rnoprRps  Kr!c(eZdZfdZfdZxZS)DebugInterpretercTt|jg||_t||yr)r modulesymbol_mappingsuperrun)selfr2 __class__s rrYzDebugInterpreter.run}s%*4;;>>  Tr!c fd fd fd fd}t|}d|jvrtj|jd\}}tj|\}}t |t |k(sJt |dt |t tt |||D]/\ }} t| tjs"||| fd1|S)Nct|ts|Stj|jj j j}|jsJ|t|Sr) r+rsympyexpandrexprxreplacerW is_numberint)nirrZs r subst_symintz/DebugInterpreter.run_node..subst_symintsSb&)  RWW\\2243F3FGHA;; ! !;q6Mr!c,tfd|DS)Nc3.K|] }|ywrrP)r.rdrfs rr0zHDebugInterpreter.run_node..subst_symint_tuple..s8Cbb)Cs)tuple)nisrfs rsubst_symint_tuplez5DebugInterpreter.run_node..subst_symint_tuples8C88 8r!c|jdkDr`t|jD]H}|j||j|k7s-|j |dkDsHyy)Nrr FT)numelrangendimstridesize)abidxrfs rcheck_significant_stridesz.check_significant_stridessgAGGI&* =C#AHHSM2ahhsmC UVU[U[\_U`HadeHe$)r!c "t|sJ|j|jk(s(J|d|jd|j|j|jk(sGJ|d|jd|jd|j||}|sGJ|d|jd|jd|jy)Nz:  != z aka )callabler)rqrp)nvrvdesc same_stridesrurks rcheckz(DebugInterpreter.run_node..checksD> !>88rxx' NDF82bhhZtBHH:)N N'%bggi0BGGI= \6("RWWYKu-? -J,K4PRPWPWPY{[ \=4Rz+DebugInterpreter.run_node..ss'$:M:M9N&Or!) rXrun_nodemetapytree tree_flattenr1ziprnr+rTensor)rZnr}ren_valsn_specr_valsr_specryrzrurrfrkr[s` @@@@rrzDebugInterpreter.run_nodes  9  s G Q  AFF?#00?NFF#003NFFv;#f+- P#f+d3v;-/P P- s6{!3VVD 2r!"ell3b"OPEr!)__name__ __module__ __qualname__rYr __classcell__)r[s@rrTrT|s**r!rTc,t|jS)z Returns a (slow) interpreter over the FX graph module that also checks various debugging properties (e.g., that tracing strides matched real strides.) )rTrYrQs r debug_noprs D ! % %%r!ct|tjj|}tjj |j }|Sr)rrr8r9r;r<)rrIrCs rsimple_ts_compilers=D A "A Hr!c"t|tSr)r r)rCs rnnc_jitrs , --r!c0t|j|Sr)rGrHrQs r print_compilers $)) Kr!fnc ttttd}|j|t |t j jr t|fi|St|fi|S)a Wrapper function over :func:`aot_function` and :func:`aot_module` to perform memory efficient fusion. It uses the :func:`min_cut_rematerialization_partition` partitioner to perform efficient recomputation. It uses NVFuser to compile the generated forward and backward graphs. .. warning:: This API is experimental and likely to change. Args: fn (Union[Callable, nn.Module]): A Python function or a ``nn.Module`` that takes one ore more arguments. Must return one or more Tensors. **kwargs: Any other overrides you want to make to the settings Returns: Returns a ``Callable`` or ``nn.Module`` that retains the eager behavior of the original :attr:`fn`, but whose forward and backward graphs have gone through recomputation optimizations, and the graphs have been compiled with nvfuser.  fw_compiler bw_compiler partition_fndecompositions) rDrdefault_decompositionsupdater+rnnModuler r )rr3configs rmemory_efficient_fusionrsW6"!;0 F  MM&"ehhoo&"'''B)&))r!c|jdtd|Dcgc]}|j|jfc}dddlm}|j |t||Scc}w)NfooaQ ############################################################## # To minimize FX graph, copy and paste the below and run it # ############################################################## import torch import torch.fx as fx from functorch.compile import minifier, check_nvfuser_subprocess, check_nvfuser_correctness_subprocess inps = a? inps = [torch.ones(shape, dtype=dtype, device='cuda') for (shape, dtype) in inps] from foo import FxModule mod = FxModule().cuda() with torch.jit.fuser("fuser2"): # check_nvfuser_subprocess can be replaced with check_nvfuser_correctness_subprocess minifier(fx.symbolic_trace(mod), inps, check_nvfuser_subprocess) r)FxModule) to_folderrGshaper)rrcudarD)rr?rrs r debug_compiler ssNN5  &**T!''177 T*+, (HJOOt dD !! +sA/ c Jg}t|d5}tj|}g}|D]}t|dk(r|}|t j }n|\}}}} } | t jt jt jt jt jt jtthvrt jdd|| | }nt j || | }|j| ddd|S#1swY|SxYw)zZ Return a random input for the given inputs meta generated from _save_fx_default. rbr r)r)r5N)openpickleloadr1randomrandrrcint32int64booluint8floatrandintappend) input_data_pathinputsrC inputs_metarr6inputrrpr)r5s r get_inputsr-sF  %!kk!n D4yA~V[[]+592eVUFIIKKKKJJIIKK  "MM!QU6RE!JJuE&IE MM% '  &. M/ &. Ms C>DD"ct ddlm} fd  fd fd} fd} fd}||||||tS) aO The forward, backward, and joint computation graph will be stored in {folder_name}/{current_name}/{current_name}_forward_{graph_index}, {folder_name}/{current_name}/{current_name}_backward_{graph_index}, and {folder_name}/{current_name}/{current_name}_joint_{graph_index} respectively. The input shape of the graphs will be stored in the .input files. These files can be loaded with pickle, and is a list of format (type, shape, stride, dtype, device). In the case of type = int or float, it is just (type,). For joint graph input, it is a nested list [[],[]] where the two inner lists have the same format. If dump_example_input is True, example_inputs will be stored in .pt file. Since each function might produce multiple graphs, the graph_index is used to distinguish difference graphs r)aot_module_simplifiedcg}t|dkDr1t|dtr||dz }||dz }|S|D]}t|tk(st|t k(r|j t|fC|j t||j|j|j|jf|S)Nrr ) r1r+rir6rcrrrrpr)r5)r2 input_metaargget_input_metas rrz(_save_fx_default..get_input_meta^s t9q=ZQ7 .a1 1J .a1 1J CCyC49#5!!49,/!!#Y 3::<CJJO r!c6t|jjdk(r,tjtj d|t ytj|}|jjtjjj|j |}tjj!d}|stj"d|j%ddd|dt t'j(|t+ddd|dt dd|dt ddr7tj,|ddd|dt dd|dt dyy)Nrz!No nodes in graph {%s}_{%s}_{%s}./rIz.inputwbz.pt)r1rrlogloggingWARNING graph_indexcopydeepcopy set_codegenrfxCodeGenrospathexistsmakedirsrrdumprsave) gm_to_saver2 type_namegmrisExist current_namedump_example_input folder_namers rgraph_saver_helperz,_save_fx_default..graph_saver_helperms z%% &! + GG3    ]]: & UXX^^3356 #D) ''..K=,!@A KK;-q7 8 m1\N!L>9+Q{m T    -qa ~Qyk;-WXYeXffghqgrrsttAAGH    JJ-qa ~Qyk;-WXYeXffghqgrrsttAADE  r!c||d|S)NforwardrP)rfw_argsrs rgraph_saver_forwardz-_save_fx_default..graph_saver_forwards2w 2 r!c.||dtdz a|S)Nbackwardr )r)rbw_argsrs rgraph_saver_backwardz._save_fx_default..graph_saver_backwards2w 3q  r!c0||dt||S)Njoint)r)r joint_argsrs rgraph_saver_jointz+_save_fx_default..graph_saver_joints2z73 Z00r!r)functorch.compilerr) rrrrexample_inputsrrrrrrs ``` @@r_save_fx_defaultrLsC 8 #J 1 ! '(&-  r!c*datt|||S)as Dump the forward, backward, and joint computation graph. Example Usage: save_fx_func = graph_dumper_aot(current_name, folder_name, dump_example_input = False) optimize_ctx = torchdynamo.optimize( save_fx_func ) with torch.enable_grad(): with optimize_ctx: result = forward_and_backward_pass(model, example_inputs) r)rrr)rrrs rgraph_dumper_aotrsK #\;@R SSr!)T)F)Vrrrrr contextlibr functoolsrtypingrrr^rrtorch.fxrtorch.nnr torch._decompr%torch.fx.experimental.symbolic_shapesr aot_autogradr r r compile_utilsr partitionersrrrtorch.utils._pytreeutils_pytreer getLoggerrrr r& GraphModulerDrKrNrR InterpreterrTrrrrrdetach gelu_backwardleaky_relu_backwardsigmoid_backwardthreshold_backwardhardtanh_backwardhardsigmoid_backwardhardswish_backward tanh_backward silu_backward elu_backwardcudnn_batch_normcudnn_batch_norm_backward masked_fillScalarrelu leaky_reluhardtanh hardswish hardsigmoid conj_physical is_same_sizerrrrrrrrrrPr!rrs\  %" ,>GG* %$g! ??. R^^. h. . b   bnn H  /r~~/d&BNN&(&&  . yy~~KK""HHOOMMNN-2,,BC $* h !"$*N": >[~Tr!